How can you use notebooks in Fabric for data science work with lakehouse data?

• A. Notebooks (Python/R) can read from lakehouse tables, perform data exploration, feature engineering, and model development, and instrument experiments with versioning.
• B. Notebooks cannot access lakehouse data.
• C. Notebooks are only for visualization tasks.
• D. Notebooks only run on local files, not lakehouse data.

Answer: (A)

Notebooks in Fabric act as an integrated data science workspace that can connect directly to lakehouse data, so you can treat lakehouse tables as native data sources in Python or R. You can read those tables into your analysis environment, perform interactive data exploration, visualize patterns, and compute statistics to understand the data. From there, feature engineering becomes natural: you can derive new features, join additional data, and prepare datasets for modeling, all while staying connected to the lakehouse rather than exporting to separate files. When you move to model development, you can train and evaluate models within the same notebook using the lakehouse data, keeping data governance and data freshness aligned with your workflows. For experimentation, Fabric supports tracking runs with parameters and metrics, enabling versioned experiments and reproducibility of results. This end-to-end capability—reading lakehouse data, exploring and engineering features, building models, and versioned experimentation—shows why notebooks are the right fit for data science work with lakehouse data.